 Good morning from VMWare Explorer day three on the floor here at the Venetian Expo. Lisa Martin with Rob Stricci. We have been covering VMWare Explorer. This is our third day of coverage. A lot of content here. We've had some great discussions. We always do with VMWare executives, partners, customers, analysts, influencers. We're going to be talking to IBM next about containers and VMs living in harmony. Pete Brie is back with us, Global Product Executive at IBM. Bradley Knapp is here as well, Product Leader IBM Cloud for SAP at IBM. Guys, welcome. Good morning. Good morning. Thanks so much for having us. Pete, you had a talk at VMWare Explorer this week about containers and VMs living in harmony. How does that happen? What did you talk about? Excellent question. Given my experience at Red Hat and what we saw in terms of the trends. I still believe we're at the beginning of a great evolution in the industry. People are dealing with these challenges around hybrid cloud and how do I bring AI in and generative AI as the latest thing. The challenge that they have is how do we move from yesterday's architectures and the way things were done into this brave new world of containers. What I'm finding here on the show floor and even talking with our various different customers is that they really need flexibility is really what it's about at the end of the day. Let's face it, virtual machine technology is not going to go away anytime soon. It's here to stay for a long, long time, but people are attracted to modernizing their applications, modernizing their infrastructure. As they go about that journey, they're looking for ways that they can do that, maintain their current environments, but move to these new environments and get a lot of the benefits of those new environments. We've been working between IBM and Red Hat very closely to recognize that that's a need in the marketplace, but also deliver some of the technology that's going to enable that to happen for them. What are some of the, give us maybe an example of a customer who's going through that modernization. Where do they start? What's that process like and how does IBM a facilitator? Yeah, in this particular case, it's a healthcare software company. I actually just spoke to them this morning and they were talking about how they've already modernized a part of their portfolio and they have a very vast portfolio of software. They've modernized part of it and they've been working very closely with Red Hat to port their applications onto OpenShift, but they still have some of the interfaces and some of the APIs are still in their legacy environments. And so the conversation they were having with us is, how do we do this? What tools do you have available to help us assess the applications that we have today so that we can make some wise decisions and some wise choices as we move down this path of modernization? Yeah, and it would seem like this is a journey that people go both in the cloud and on-premise as well. On-premise is not going away, especially for newer applications. I think we see less quote-unquote repatriation, but more, hey, net new, we're going to start here and then maybe we take it to a cloud. And it would seem that VMware is kind of the glue between cloud and on-prem for you as well. Is that what you would... Oh, I think that's absolutely a fair statement. So if you think about the challenge in a company that's running a really large estate, right? So they've got a large on-premise hosted colo, whatever it may be, but their own estate that they own and run. And then there need to also be flexible to get workloads into the cloud, maybe to stay, maybe just for a little while. You have to have some sort of an underlying orchestration technology that makes that happen, right? VMware makes it very easy for that hybrid cloud scenario to happen where it's not necessarily repatriation, it's putting the workload where it makes the most sense. If you have something that's really peaky bursty that needs the ability to scale up and down quickly, that's the kind of thing that's perfect to go out into the cloud. Always has been. If it's the kind of thing that's very steady state, it's going to run forever, it's really not going to see peak burst, all that. Maybe that makes sense to put on-premise because you look at a TCO calculation and you say, man, the numbers just work a whole lot better. I can run this on some older hardware. It's got the latest version of ESX on it. It'll be just fine in there. But it's hard to know in the new product development cycle which one's it going to be, right? Sometimes you know, sometimes you don't. And so the ability to port your VMware workloads, run them on-premise, run them in cloud, run them in a hybrid manner, gives you that flexibility that you need that you're not going to get if you don't have the glue that stitches the hybrid cloud together. Yeah. And I think a key aspect of that is not just the application side of the equation, but the data that goes with the applications. And it's interesting, we were talking with Dave yesterday about this concept of entropy in the industry right now. And there's a lot of change happening. And people are looking for, what's my landing spot? What can I count on here? And the ability to move the data together with the applications where we're talking about on-prem, VMs to containers, containers to VMs, hybrid cloud, doing that on a global basis. There's some really exciting technology that I think about it as solving the speed of light problem, right? Because moving large data sets, moving applications is one thing, but moving very large data sets is a whole different problem. And there's newer approaches to being able to solve that problem in a hybrid, multi-cloud context. From a multi-cloud perspective, we talk so often about so many companies are in the multi-cloud situation kind of by, well, whether it's M&A, Shadow IT, whatnot, it's complex. And we heard from the show this week, and we hear this from a lot of companies, what they're doing to help customers dial down the complexity and really turn their multi-cloud environment into a strategic advantage. How does IBM help customers achieve that as they're living in this hybrid multi-cloud world? Yeah, that's an excellent point. And one of the things that we've been working very closely with Red Hat and Red Hat OpenShift in particular is a product called IBM Fusion, which is really, it's available as a hyper-converged platform, but it's also available as software because, again, back to the flexibility point, I might be running on the public cloud, I might be running on-prem, and I want a turnkey hyper-converged system that I can just drop in and fire up very, very quickly. And so we've built in, and a lot of this is obviously based upon the concepts of Kubernetes and the degree of automation that Kubernetes delivers through OpenShift, things like operators, and really automating as much of that because I think that's the new mantra in this new world, is anything that can be automated should be automated because that's the only way we're going to be able to deal with the scale and also deal with the entropy and the level of change that's happening, too. Yeah, and I would say that it seems like that bringing that type of technology to your point, I mean, data, I always say data has weight, it has gravity, and it tends to be, it would seem like, again, you want to bring, sometimes you want to bring the apps to the data versus bringing the data to the apps or where have you. How does that play with Fusion and what you guys are putting out there now? Yeah, that's a great question, and we have, one of the reasons I actually came to IBM a year ago is actually because of this technology. And I'm a storage guy, you are too, you are also, and we have this acceleration technology, and many of you have probably heard about WatsonX and our play for generative AI, and Fusion is a key part of WatsonX, there's this thing called WatsonX.data, but we have this acceleration technology that can actually cache data globally around the world, so we can go find the data sets and we can bring it to the applications, and we can solve this problem, you know, the speed of light problem that I was talking about earlier. Yeah, I mean, I think that seems exciting because, I mean, again, I know, we talked a little bit beforehand about SAP and the modern apps, bringing modern apps to something like a heritage app, like SAP, still people run a lot of mission critical stuff on, you know, manufacturing. I knew I was in the oil and gas industry, and we had a lot of SAP there, and I know because I have, like, my head goes back to actually writing a BAP code and all of that stuff for it, but how does that, how does the Fusion and what's going on in cloud and that data and WatsonX all come together? Well, so if you think about an ERP estate, just at kind of a high level, right, an ERP estate is not terribly exciting, but it's the most important thing you have. It's all of your invoicing, all of your PO, all of your inventory, what's going to whom, where, how, and when, right? These are not exciting things. It's programmatic issuing of trouble tickets so that you can fix something before it breaks, but the fusion of the data and AI components into the ERP estate, that's the really interesting part, right? The cool thing in AI right now, everybody wants to talk about LLMs. Oh, we got to go talk about the LLMs. LLMs are the greatest thing in the world. It's kind of fun to have a conversation with a computer, right? But in the world of enterprise, don't get me wrong, LLMs have very distinct places, but the large generative AI can do so much more if I am in the manufacturing sector, is it more useful for me for my, you know, 50 customers to be able to have a generative AI that's an LLM that they can ask, like, hey, when's my next shipment going to be here? Like, that's kind of cool, but is that anywhere near as useful as feeding hundreds of thousands of sensors and all of that edge data, bringing it back into a central processing location, feeding that through the generative AI, and then all of a sudden figuring out, okay, we have no idea why, but when this, this, and this combines, when we get these three anomalous inputs, that's going to lead to this fourth very bad thing. We cannot explain why. The computer's the one who told us how, but when this happens, we know that we need to go out and fix it. That's not cool unless you're in the industry, but if you can get that programmatic ticket to go out and repair before something goes wrong, maybe that stops a 50 or 100 million dollar outage or a 30 million dollar crash on a machine or a tool and die crash where you've then got to be down for weeks before new tools can be built. Like, that's the genuinely exciting things. Now, it's exciting in a very weird way. It's a tech way exciting. It's not having a chat with a computer, but at the same time, the fact that you already have all of this data that's spread all throughout the world and maybe you can have, going back into manufacturing again, maybe you can get learning from sensor platforms in, I don't know, Australia, that are going to apply to your manufacturing location in Germany. And the ability to bring all that edge data back to cache it so that your machine learning algorithms can work their way through it, make those predictions and you can look at it and you can say, I've no explanation for why this is, but now we know that it is. The machine proved it out. It showed us its work and holy cow, now we're off to the races. Definitely. And I think what's interesting is that IBM has a full stack here and I think this is why I get excited every time I talk to you guys because I think it's one of these things that I learn something new every time about what you've been doing and I've known Watson for, you know, even before it got rebranded and what the way it's going and LLMs blew up and you know, I tried to use it with weather channel data at one point in time to do something with our product that I was running. And I think what's interesting is that there's a, I mean, data is that you have to have a data lake and but it's also you want to keep it secure and to your point on LLMs, I want, maybe it's not an LLM or it's an SLM like a specific or segmented large language model type of thing. Is that what you guys are bringing together with some of the fusion stuff and data lake stuff that you're doing? Yeah, absolutely. You know, we're working with the large language models but we also recognize, particularly as you look across, you know, vertical industries, you know, those models aren't necessarily, they're going to be more specific for those industries. You know, so we're doing that kind of work. We've built a data lake house, you know, that works hand in hand with Watson X, you know, so you can get the advantage of, you know, tapping into a lot of the new formats, a lot of the parquet formats that people are now using and be able to do selective queries. And, you know, I mentioned the acceleration technology earlier and some of our early, you know, research that we've done, we're able to speed up queries easily, orders of magnitude. So it's really exciting to us and again, it just underscores why it came to IBM. You're absolutely right. I mean, there's a vast portfolio of technology we have and I think IBM, you know, the new IBM, we're really figuring out how to align that technology to solve a lot of these business problems. No, yeah. Go, sorry, just double click into some of those business problems and really highlight from a fusion perspective, what are some of the key benefits it's delivering for customers? You talked about faster query speed, I think, performance, from a high level business perspective, what are some of the impacts that you're able to make? Go ahead and tell you what, let's jump into the reality of business use of AI for a moment, right? So we've got a brand new product out, the Watson X Studio, where the idea behind Studio is to make it possible for people who are not PhD data scientists to teach and more importantly, evaluate the effectiveness of different AI products. They could be LLMs, they could be SLMs, because IBM, we realize we are not going to write all of the world's best AI products. It's not going to happen. Now, we've got some really good ones, we've got some world leaders, but everything is going to be IBM and so the idea is to give a non-data scientist a single location to come in and they can say, all right, look, I want to evaluate these four models against each other to solve a very specific problem. My very specific problem is I have 50 years of contracts, and with amendments and statements of work and all of that, and I want an AI solution that is going to read my 50 years of contracts and I want to be able to ask it, synthesize for me a single contract that shows everything that's enforced for customer X, right? That's a really cool problem to solve. You could pay a hundred people to work a thousand hours each and do that, or you can have the AI model do it. But where it gets really interesting is, okay, there are maybe five or six different SLMs that you want to evaluate on how to do that. So the first thing you've got to do is you've got to train them all and then once you've trained them all, you've got to evaluate their efficacy. Well, again, you could hire some very expensive data science PhDs or using a product like Watson Studio, Watson X Studio, you can have somebody who is good at data and maybe good at technology and understands AI but doesn't have a PhD in data science because you can't be hamstrung by the inability to access a world expert to solve that business problem. You just need to know what contracts apply to that customer. It's like I talk about the democratization of these technologies and making it available to everybody in these organizations, not just the data scientist or not somebody who's peaked who has a PhD. The other big important piece that I want to get in there because again, you've got two guys from IBM and so you're never going to get away from a conversation with us without talking about Watson in some way, shape, or form, but also about trust and governance. I think everybody has seen on the news how AIs can go rather badly wrong if the prompt engineering is done correctly and things like that. That is certainly one part of trust and governance. The second part of trust and governance is being sure that your AI doesn't hallucinate and just start making things up. The third part is really around securing the data that you use to train your AI but also ensuring that I don't want my data to train an AI that's used by my competitor and I want to know that my results are confidential that I am not improving the outcomes for somebody else. So governance is not just about people, it's also about process, it's about security because when you go in and you have these enterprise conversations, if I go and I ask chat GPT and I don't know, tell me how to mix a paint, that's not exciting stuff but if I'm feeding all my contracts into an AI, I don't want to improve the contract reading ability of that AI for my competition because what happens if somebody can prompt engineer their way into my contract data? That's terrifying. And so that underlying trust, that underlying governance is so, so, so important and that's one of the things that I think all of the enterprises are looking at in this new exciting world of AI that they've got people that are pushing them, go, go, go and it's like, all right, hang on. We need to be able to have trust and faith that the decisions that the machines are suggesting are the right decisions but also that we're not going to accidentally send our confidential info out into the world. Makes total sense. It does. I'm just like, from a security perspective, you bring up such a great point. I'm just curious, when customers come to IBM talking about wanting to work with you guys to build AI products, what does a typical road map look like? Because, you know, you hear from executives around the world, if you're not in AI yet, you're behind. But what does a journey look like? I imagine it... So it is a journey. It absolutely is a journey, right? So in this, you asked one of my favorite questions because every, you know, AI is hot topic, right? Everybody wants to talk about it. And so it's go, go, go, go. I was actually presenting at an AI conference a couple of weeks ago and as part of my session, I asked everyone who was there what's the first step in an AI journey if you're going to go out and you're going to engage some consultants and engage somebody like IBM to help you out. What's the first step? And I got all kinds of answers. But none of the people gave the answer that is my personal favorite which is what business outcome am I trying to achieve? Exactly. By implementing AI. Exactly. And there were like, there were maybe two people that they're like, yeah, we want to start with that. So the first piece of an engagement is actually figuring out what is it that we're trying to achieve? The second piece, what are the metrics that show success? And it's amazing how often people want to skip those two steps and they want to go right into, let's go and evaluate some models. And it's like, hang on. Well, tech is cool, right? It is. That part of it. Get into it. And I think that's a big piece of it is that, and I totally agree. It's like, what is the ROI that I'm going to expect to get? What is the business problem that I'm solving before throwing AI at it? Because you may actually come up with unintended consequences from doing that. Absolutely. So definitely a very interesting piece of it from that perspective. Yeah. But so, oh, sorry. No, go ahead. I think what kind of bringing it back around, any other announcements or anything that you wanted to get across that we didn't cover yet today? No, I think, I mean, the big news that came from us recently is the WatsonX announcement and the WatsonX.ai, which is the core product. And then we delivered the data lake house to go along with it, which is WatsonX.data, which consists of IBM Fusion together with IBM CEF technology for the large object store capabilities and then there's also WatsonX.gov for that specialized market. So, yeah. That's all good stuff. Always exciting. And for anybody who's out there that just wants to go and play with it, there are some of those solutions that do have costs associated with them, but they don't all, right? Anybody can go out and spin up a trial account with the WatsonX Studio product to go and at least kick the tires a little bit, see what it's all about, see what the purpose is. I have one, right? I have a free trial account that I go and I play with because it genuinely is kind of neat to do ABCD testing of four different products. You know, I'll have like an IBM product and then three other products that all say they do the same thing. And, you know, I like to feed in the same training data set and see what the results are. And, you know, that's just because I'm a big nerd. Sometimes I have free time, not so much this week, but other weeks. It's just, it's kind of fun. So I want to encourage folks to go out and, you know, kick the tires a little bit, poke it and see what you think. It's really interesting at IBM and this tells you how much our leaders believe in, you know, the direction that this is headed is every IBMer had to participate in a WatsonX challenge. So it didn't matter if you were techies like us in product management or engineer or in procurement or legal or HR, you all had to participate in the challenge and you had to get hands on time with WatsonX. Wow, cool. Pete, take us out here with kind of, you know, global product executive, peek into the road map. Some of the things you talked about, some of the news, what's going on, but what can we expect, say, in the next six to 12 months from IBM? Anything you can share? Yeah. And I think it goes along the lines of what we were talking about earlier, you know, the democratization, but also the key part of that is making it simpler to consume technology. And so you're going to see some key technology coming from Red Hat that we're working very closely with them, you know, to be able to manage large scale clusters. That's going to be a key piece of technology. You're going to see continued improvements in this acceleration technology because we realize you know, at the end of the day, these organizations are looking for productivity improvements. So those are going to be two key areas that we're going to be pushing on in the next six months. All right. We're going to keep our eyes on this space. Bradley, Pete, thank you so much for joining Rob and me on the program today and sharing what's new at IBM, some of the exciting things in hybrid, the hybrid story being successful in AI. We're definitely going to watch this space. Thank you guys. Our pleasure. Thank you. For our guests and for Rob's Stretch A, I'm Lisa Martin. You're watching theCUBE live from day three of VMware Explorer. We're going to be back on the other set with more great content. So stick around. You're watching theCUBE, the leader in live tech coverage.